In communication systems which transmit data over a wide band of frequencies, it is often useful to segment the frequency band into multiple sub-channels (or tones). Each of these sub-channels carries a portion of the total information sent from the source device to the destination device.
The proportion of data transmitted on each of the different sub-channels is often configured based on factors such as the data carrying capacity of the sub-channel. The data carrying capacity of a sub-channel is related to its bandwidth and the physical signaling conditions of the sub-channel.
In communications systems such as power line communication systems and others, it is desirable to maximize the throughput on the set of sub-channels allocated to a connection between two devices. In order to maximize the amount of information a sub-channel can carry, the system must characterize the transfer function of the channel by making measurements of the channel. These measurements may include, but are not limited to, such attributes as signal to noise ratio (SNR), channel error rates (e.g. bit error rates, symbol error rates, packet error rates), phase noise and adjacent channel interference.
In order to take measurements on the sub-channels, the communication system may allocate a percentage of the transmission time or other network resources for this purpose. The channel measurements thus may have a negative effect (reduced bandwidth) on the ability of the channel to carry data. However, it is expected that the net gain in channel bandwidth is positive due to the increased performance obtained from the channel by using the measurements to maximize the throughput of the individual sub-channels.